Analysis of the impact of the COVID-19 pandemic on the stage at diagnosis in breast cancer patients at a French comprehensive cancer centre, through two different methods: a preliminary study.

Abstract

Purpose: In early 2020, the Coronavirus-19 (COVID-19) pandemic led to widespread lockdowns, disrupting cancer-screening programs and limiting access to care. Although a temporary drop in new breast cancer diagnosis had been noted, variations in stage of disease have been explored less frequently, and with methodological approaches that might lead to imprecise or approximative results. This preliminary study aimed to assess possible variations in breast cancer stage at diagnosis over a long-time period using two different approaches.

Methods: We analysed data from 3 787 women with invasive breast cancer treated at our comprehensive cancer centre between 2017 and 2022. We evaluated changes in proportions of staging parameters using two different approaches: a traditional "pre-to-post pandemic" traditional comparison and time series models. The latter included ARIMA (AutoRegressive Integrated Moving Average) models, complemented by the research of potential significant estimated structural breakpoints over time in linear regression models.

Results: The pre-to-post comparison suggested an overall positive overview of the differences observed before and after the pandemic. However, ARIMA and logistic models demonstrated a relative stability in tumour size and metastatic status, with only one significant breakpoint observed: a shift in the rate of patients with no lymph node involvement (N0), likely unrelated to the pandemic.

Conclusions: This is the first study to assess changes in breast cancer stage at diagnosis using time series and structural breakpoint analysis over an extended period. Our preliminary results highlight the importance of using advanced statistical techniques when evaluating the impact of systemic disruptions (like COVID-19) on cancer care.